摘要
为解决时变多径信道中调制分类问题,提出对数字通信信号先分段盲均衡,后星座图匹配的识别策略,保证了每段信号对应的信道环境近似静态以及降低码元速率和载波频率的估计误差对均衡的影响。选取与信号星座图形状无关的分数阶盲均衡算法,以满足调制识别过程中均衡算法对调制类型盲的要求。信道时变使得各段序列均衡效果表现很大差异,采用极大似然比方法挑选最佳均衡段。对最佳均衡段通过星座图匹配实现调制类型判别。仿真结果表明:在典型城市和恶劣城市模型下,信噪比大于5 dB时,平均正确率在90%以上,该算法对时变多径信道非常有针对性。
Segmented blind equalization and constellation matching is used for signal classification in time-variant multipath channels. The Segmentation seeks to ensure a static channel environment in each segment and to relieve the influence on the equalizer of the carrier frequency and the symbol rate estimation error. The Fractionally Spaced equalization algorithm was used to the blind recognition. The equalization result in each segment differs because of the influence of the time-variant channel. The optimal equalized sequences are selected based on the maximum likelihood function with the modulation type selected based on constellation matching of the optimal sequence. Simulation results show that the average success rates are more than 90% in typical urban （TU） and bad urban （BU） channels when the SNR is above 5 dB.
出处
《清华大学学报（自然科学版）》
EI
CAS
CSCD
北大核心
2007年第7期1097-1100,共4页
Journal of Tsinghua University(Science and Technology)
关键词
调制识别
时变多径信道
似然函数
modulation recognition
time-variant multipath channel likelihood function
